Gemini 3.8 Flash TTS vs Sume TTS: break-even is 47 characters
Gemini 3.8 Flash TTS costs $0.00225 per 10 seconds of audio; Sume TTS costs $0.0475 per 1,000 characters. They tie at about 47 characters per 10 seconds.

Google bills Gemini 3.8 Flash TTS by audio length: $0.00225 per 10 seconds through December 31, 2026. Sume bills TTS by characters: $0.0475 per 1,000. The two meet at about 47 characters of script per 10 seconds of audio ($0.00225 / $0.0000475 = 47.4). Real narration usually packs more characters than that into 10 seconds, which would put Google ahead, but that depends on your voice, language and pacing, so measure your own script.
The break-even line
Because one is priced per second and the other per character, the speech rate decides the winner. The table gives the cost of 10 seconds of audio at three script densities. The densities are hypothetical examples, not measurements of any voice.
Cost of 10 seconds of audio at three assumed densities.
| Characters spoken in 10 s | Sume at $0.0475 per 1,000 | Gemini 3.8 Flash through 2026 | Gemini from 2027 |
|---|---|---|---|
| 30 | $0.00143 | $0.00225 | $0.0045 |
| 47.4 (break-even, 2026 rate) | $0.00225 | $0.00225 | $0.0045 |
| 95 (break-even, 2027 rate) | $0.00451 | $0.00225 | $0.0045 |
| 150 | $0.00713 | $0.00225 | $0.0045 |
The January 2027 change
Google's page lists the audio output rate as $9.00 per million tokens through December 31, 2026 and $18.00 starting January 1, 2027, so the per-10-second figure doubles to $0.0045. The break-even then moves to about 95 characters per 10 seconds ($0.0045 / $0.0000475 = 94.7). A script that is cheaper on Sume in 2027 is one that speaks fewer than about 95 characters per 10 seconds.
Gemini 3.8 Flash also has a batch tier at $4.50 per million audio tokens through 2026, about $0.001125 per 10 seconds, which moves the 2026 break-even down to roughly 24 characters.
Constraints on the Sume side
A single Sume TTS request is capped at 20,000 characters, which costs $0.95. The catalog lists sume/tts-1.0 and sume/tts-router at the same character rate. Sume's rate is easy to forecast from the script length alone, before any audio exists, which is the practical advantage of per-character billing.
Sources
Related posts
More in Comparisons
- Laugh and sigh tags in TTS: Gemini 3.8 tags vs Sume's emotion field
Gemini 3.8 TTS takes inline tags such as laugh, sigh and breath. Sume TTS documents an emotion hint, speed and volume instead; test tags on one short job.
- Generate at 1K and upscale, or generate 4K? Sume price check
Nano Banana 2.1 at 1K plus a Sume image upscale is $0.30, against $0.20 for native 4K. GPT Image 2.5 low plus upscale is $0.22 against $0.22 for high 4K.
- Griffin-Lite's 0.43 s latency: what a recorded avatar clip gives up
Tavus reports Griffin-Lite video latency of 0.43 s on average (0.27-0.59 s) in a research preview. A recorded clip on Sume is a job, so choose by use case.
- HappyHorse 1.1 lists 9 aspect ratios; Omni on Sume lists 2
HappyHorse 1.1 reference-to-video offers nine ratios, from 9:21 to 21:9. Gemini Omni Flash offers 16:9 and 9:16. How to check a Sume model's ratios first.
Written by Sume